Star trailing

[Stub Mandrel]

I'm getting start trailing in RA over even quite short exposures. I spent an hour checking my mount was being driven at the right rate and over an hour the RA was being driven at the right rate to within 1 second, but even 30-second subs show trailing - I'm getting about 2 or 3 decent subs out of every 12. It isn't a lack of rigidity as there is no blurring or movement in dec at all (once polar aligned properly). I'm not expecting miracles but I am sure that unguided 2-minute subs ought to be possible with my kit (NEQ3 on a EQ5 tripod).

I have some clues:

1. When I first tried out the gear, I got better results but I have 'fiddled' (with just about everything) - so it probably is something I have misadjusted.
2. Sometimes instead of a trail I get two 'spots' for each star, or a 'tadpole' as if its moved then settled down (or the other way round) - this suggests the drive isn't smooth enough and its jumping.
3. Sometimes a meridian flip helps (or makes it worse) - this suggests it might be backlash.
4. After changing target it takes 2 or 3 minutes to settle down - again suggests backlash or lack of smoothness.
5. The length of the trails is fairly short, but not consistent. they get longer with longer exposures which suggests they aren't periodic error.

Here's an example (single sub unstacked and partially processed):

One of my main 'fiddles' has been to apply teflon grease to the worm and wheels, and also to try and remove the backlash. I THINK that I have made the drive to stiff in trying to get rid of backlash, instead of just using the weights to provide slight east-bias.

There seem to be so few cloudless nights trial and error is a very slow way to get it right :-(

It would be really helpful if someone could  give me an idea of how best to balance the stiffness/freeness of the RA drive with the amount of backlash.

[jambouk]

What sort of scope is it and what is its focal length? Are you using an coma correctors / flatteners?

How do you know it's not imperfect polar alignment?

James

[ronin]

Your signature say 150PL and NEQ3, so more information or clarification would be useful.

If the scope used was the 150PL I think you will be lucky to get an exposure of the length you are suggesting. The 150PL is 1200mm long and the scope itself has a fair cross section so any wind or breeze will be a problem. In general imaging scopes are a fair bit smaller. I would thought that even the 130PDS would be a boarderline and the WO ZS 71 a better item.

If you are imaging with a DSLR and a 150PL then that is a fair bit of weight for accurate tracking on what is an EQ3 head. The payload for an EQ3 is 5Kg. I would guess you are over that and if you throw in the long focal length and the side area you likely do not have a stable imaging arrangement.

Gears will always have to have some play, they are simple mechanical items and not even precision ones. Synta will not go to that extent. So trying to reduce the play to 0 could mean they re too tight.

I assume you have a polar scope for the mount.

[DaveS]

Agree with ronin here.

An EQ3 is marginal for AP. Yes, it can be done but takes hard work, and using a 150 PL is not making things easier. I would suggest using the mount with a DSLR lens on the camera, perhaps no more than 200 - 300 mm and limiting the length of your subs.

[jambouk]

Yes, 150 newt and dslr will be a weight for the poor mount. You might get lucky and get one OK sub every 5 or so, but that is a lot of wasted subs.

Once you've tightened up all the mechanical issues and polar alignment, i'd aim for shorter subs in the region of 30-60 seconds and just do many more of them.

Watch what you are resting the tripod on too; slabs can move, and concrete can transmit vibrations. As above, if the wind is above 10mph or transient gusta you'll also get some movement artefact.

Make sure you balance witht the scope loserd with the dslr on board, and any dew shields you use, your mount needs all the help it can get.

Again as said above, if you want longer subs with the current kit, just mount the dslr directly on to the mount and do some dalr widefield.

Good luck

James

[Alien 13]

I would agree with the people above the mount will with a bit of work get a max of 2 mins with a camera and lens and 45-90 seconds with a short refractor of sub 400mm FL the tadpole shapes you describe are an indication of vibration caused by wind etc.

Alan

[Tinker1947]

If you're stuck with the kit you have there are a few things to try, get the PA as good as the kit allows, close will not be good enough, balance the scope this will end up with the DSLR in a position that's not desirable getting the balance equal while its horizontal is half the battle, when it is position it  up right  twist the scope so it stays upright and doesn't free fall left or right, this generally ends up with the DSLR on top or underneath, check the power supply is stable and is towards the top end allowed, don't bother is there is a some much as a slight breeze it will move the scope.....save some pennies for a HEQ5........  

[kalasinman]

I have a working theory that some mounts have engineered in stiffness, or resistance if you like. This covers certain design defects and if removed, things may get worse.

[dph1nm]

The length of the trails is fairly short, but not consistent. they get longer with longer exposures which suggests they aren't periodic error.

Hmm - the worm period is probably around 10 mins on these mounts, so I wouldn't be surprised if it was periodic error. You are also looking at a PE amplitude of 10s of arcsec for this class of mount, which at 1200mm FL is going to show up in not too many seconds!

If you have an intervalometer, chose a bright star near zero dec and take loads of very short (1sec?) exposures every 15 secs or so for at least 10 mins. Then measure the star position and see if it looks periodic (it helps if you know which direction RA is). Another way to check (if you have a dark sky!) is to deliberately offset your PA, then take a very long (30min) single exposure (at low ISO) - star trails should show wiggles if it is PE.

NigelM

[kalasinman]

Guiding cures a multitude of problems, sometimes easier than fighting mysterious demons.

[jambouk]

"Another way to check (if you have a dark sky!) is to deliberately offset your PA, then take a very long (30min) single exposure (at low ISO) - star trails should show wiggles if it is PE."

I like this; going to try it sometime.

JD

[jambouk]

"Another way to check (if you have a dark sky!) is to deliberately offset your PA, then take a very long (30min) single exposure (at low ISO) - star trails should show wiggles if it is PE."

I guess if you don't have a really dark sky, you could just take lots or 30-60 second exposures and then stack them but don't align on a star.

JD

[Atreta]

I'm getting start trailing in RA over even quite short exposures. I spent an hour checking my mount was being driven at the right rate and over an hour the RA was being driven at the right rate to within 1 second, but even 30-second subs show trailing - I'm getting about 2 or 3 decent subs out of every 12.

i'm about to create a topic about something similar, i was trying last Friday to capture m31, did a drift alignment, went for 30s just to see how it was. 

every few pictures showed showed trailing, i.e picture 1 is ok, 2 trailed, 3 trailed, 4 ok and so on. could it be a balancing issue?

thanks

[jambouk]

If it was JUST a balancing issue every sub would suffer. But that isn't to say your balance is perfect. Balancing is one of the easiest components to sort out.

You need to divilge more about your scope and mount, and what surgface the mount is set up on. The commoner causes of intermittant star rrailinf include:

- wind (and not the baked beans variant)

- transmitted vibrations from the operator

- transmitted vibrations from the cameras shutter

- true periodic error

- unstable power supply (theoretically)

- gravitational waves distorting space-time

James

[Stub Mandrel]

A lot there. I wanted to think about the comments before replying, now even more to think about!

I thought I had explained my setup, but it's an NEQ3 mount on an EQ5 tripod. I'm setting it up on a patio, so effectively each leg is on an isolated brick on a sand base, so no big vibration issue.

The scope is a Skywatcher 150PL so long FL of 1200mm. I'm using an EOS 10D which weighs about 0.7kg.

I'm using mirror lock. PSU is a 12V battery so very stable.

I don't have a copy of Planet Waves in my Dylan collection, so I should be OK on that count.

Polar alignment is probably one of my problems I'm using a non-magnifying circle to show the 0.98 degrees, so I am sure I will get better alignment with a polarscope,

I've done tests using a 30-second exposure and the resulting streaks were horizontal suggesting there was negligible drift in DEC, but I notice that on my M31 the trails are all at 45 degrees and that each image shifts down and to the right of the previous one. It's just getting through to my brain that errors in PA will give different results depending on where the scope is pointed.

I am in a very sheltered place (a pain as it means I have very little sky, not enough low down in E or W for drift alignment to be done properly), I can recall one gust in several nights that was big enough for me to notice.

RA drive only, home made. With the 130 tooth worm gear I calculate it should take 11 minutes 2.8 seconds to turn the worm once. On test it took 66 minutes 21 seconds (instead of 66 minute 17 seconds) to make 6 turns, an error of 0.1% or an arc-second in 1 minute or ~ 1 pixel in 1 1/2 minutes - surely not enough to cause trailing?

Looking back to early shots, I took 14 30-second subs  of which only four have 'oval' stars, and I wouldn't go as far as saying trails. I haven't been able to get anywhere near this good since. I realise the longer subs I go for, the more I will have to discard, but it's clear I was doing better at the start.

I suspect I was lucky with PA first time, then when I struggled again I played with the worm drive and compounded the error.

I think I have to try again now I've slackened the worm drive a bit, get PA as good as I can and at least see if I can get back to were I was.

One concern I have is that I need to get unguided performance as good as possible before going to guiding - I don't want to waste effort making DEC drive and a new drive system for guiding if it's to no avail.

Am I setting myself a needless challenge and should I go straight for guiding?

[jambouk]

I don't know enough about DIY fitted drives to advise you, sorry.

Have a hunt around the forum to see what sort of results people are getting with similar kit.

James

[kalasinman]

A lot there. I wanted to think about the comments before replying, now even more to think about!

I thought I had explained my setup, but it's an NEQ3 mount on an EQ5 tripod. I'm setting it up on a patio, so effectively each leg is on an isolated brick on a sand base, so no big vibration issue.

The scope is a Skywatcher 150PL so long FL of 1200mm. I'm using an EOS 10D which weighs about 0.7kg.

I'm using mirror lock. PSU is a 12V battery so very stable.

I don't have a copy of Planet Waves in my Dylan collection, so I should be OK on that count.

Polar alignment is probably one of my problems I'm using a non-magnifying circle to show the 0.98 degrees, so I am sure I will get better alignment with a polarscope,

I've done tests using a 30-second exposure and the resulting streaks were horizontal suggesting there was negligible drift in DEC, but I notice that on my M31 the trails are all at 45 degrees and that each image shifts down and to the right of the previous one. It's just getting through to my brain that errors in PA will give different results depending on where the scope is pointed.

I am in a very sheltered place (a pain as it means I have very little sky, not enough low down in E or W for drift alignment to be done properly), I can recall one gust in several nights that was big enough for me to notice.

RA drive only, home made. With the 130 tooth worm gear I calculate it should take 11 minutes 2.8 seconds to turn the worm once. On test it took 66 minutes 21 seconds (instead of 66 minute 17 seconds) to make 6 turns, an error of 0.1% or an arc-second in 1 minute or ~ 1 pixel in 1 1/2 minutes - surely not enough to cause trailing?

Looking back to early shots, I took 14 30-second subs  of which only four have 'oval' stars, and I wouldn't go as far as saying trails. I haven't been able to get anywhere near this good since. I realise the longer subs I go for, the more I will have to discard, but it's clear I was doing better at the start.

I suspect I was lucky with PA first time, then when I struggled again I played with the worm drive and compounded the error.

I think I have to try again now I've slackened the worm drive a bit, get PA as good as I can and at least see if I can get back to were I was.

One concern I have is that I need to get unguided performance as good as possible before going to guiding - I don't want to waste effort making DEC drive and a new drive system for guiding if it's to no avail.

Am I setting myself a needless challenge and should I go straight for guiding?

As to the final question, yes you are/yes you should. I struggled for months with unguided. I did ok at times, but it took a lot of effort and still discarded no less than 30%.

Horizontal displacement is DEC error.

You could polar align with plate solving and get it right.  an example 

[Stub Mandrel]

Great, well I found a suitable bipolar stepper last night so DEC drive here I come :-)

... and I thought plate solving was what they did the morning after in a Greek restaurant.